Short-time deep-inspiratory breath-hold pulmonary perfusion SPECT/CT in interstitial pneumonia: a vascular dimension that stratifies risk independently of the GAP score

PURPOSE: Risk in interstitial pneumonia (IP) is assessed routinely by the Gender-Age-Physiology (GAP) score from pulmonary function and the tricuspid regurgitation pressure gradient (TRPG) on echocardiography. This study evaluated whether the percent perfusion defect (%PPD) on short-time deep-inspiratory breath-hold (STDIBH) pulmonary perfusion SPECT/CT carries prognostic information that neither captures. MATERIALS AND METHODS: Ninety-one consecutive patients with IP who underwent STDIBH perfusion SPECT/CT between 2012 and 2023 were analyzed retrospectively. The primary endpoint was a composite major adverse event (MAE) of all-cause death, acute exacerbation of IP, respiratory-failure hospitalization, and initiation of home oxygen therapy. Association was assessed by Cox regression and discrimination by censoring-weighted time-dependent receiver operating characteristic analysis. RESULTS: MAE occurred in 66 of 91 patients (72.5%) over a median follow-up of 45.9 months. %PPD was associated with MAE after adjustment for the GAP score and antifibrotic therapy (adjusted hazard ratio 1.2304 per 10 percentage points, 95% CI 1.0568-1.4325; P = 0.008), and, in exploratory analyses, for TRPG and the GAP score together (1.2058, 1.0314-1.4096; P = 0.019). It showed no significant correlation with TRPG (Spearman r = 0.164; P = 0.140). Time-dependent discrimination did not differ significantly between %PPD and the GAP score at 6, 12, and 24 months, and adding %PPD to the GAP score changed the Harrell C-index from 0.692 to 0.719 (difference 0.027, 95% CI - 0.009 to 0.107). %PPD was obtained in all 91 patients and the GAP score in 68 (74.7%). CONCLUSION: In IP, %PPD measures a dimension of risk that the GAP score does not capture, and exploratory analyses suggest that echocardiography does not capture it either. %PPD is obtainable when the GAP score is not. Its discrimination did not differ significantly from that of the GAP score, and adding it to the GAP score did not significantly improve discrimination.